Stop Measuring What's Comfortable. Start Measuring What Matters

There's a false belief embedded in how a lot of companies operate: that more measurement equals more control. It doesn't. It usually just creates more noise for the team to navigate around.

We're Watching a New Cost Category Be Born in Real Time

I'll start with the problem that's closest to me right now, because it's the clearest example of broken measurement I see.

Every company is adopting AI tools at a remarkable pace. But the finance function has almost no infrastructure to see, govern, or forecast what it's actually costing. AI spend doesn't behave like any cost category teams have dealt with before — it's not headcount, not traditional software. Token costs are variable, tied to how people work, and show up directly in gross margin in ways companies never modeled when they made their AI investments. (We built AI Token Attribution specifically to give finance leaders visibility into this.)

The activity metric looks great: "80% of our team is using Copilot, we rolled out Claude to the whole org." But gross margin is quietly compressing and nobody can explain why. Nobody was watching the clarity metric: what is this actually doing to our economics?

That's the problem we're building Stackpack to solve. But it's also a perfect illustration of the measurement trap almost every company falls into.

Why This Keeps Happening

The first reason is simple: we measure what's easy to count, not what actually matters. Activity is always more measurable than impact, so teams default to it. Calls made, emails sent, features shipped. These feel like progress because they go up, but they don't tell you whether any of it is working.

The second reason is that metric-setting is a political act. Pick a number, commit to it publicly, miss it, and there's real accountability. A lot of organizations have a deeply uncomfortable relationship with that. So they stay vague on purpose. Fuzzy goals, target ranges, nobody ever actually wrong. I've done it myself. It's institutional self-protection.

The third reason — and I think the most underappreciated — is Goodhart's Law — the idea that once a measure becomes a target, it ceases to be a good measure. Once people know what you're tracking, they optimize for the metric rather than the outcome the metric was supposed to represent. Customer satisfaction surveys get sent at the peak of happiness. Sales metrics get hit through behavior that quietly erodes margins. The best leaders understand this, rotate their metrics, and regularly ask whether the number still means what it's supposed to mean.

And then there's what I see specifically in fast-growing companies: the metrics that got you here aren't the ones that get you to the next stage. Companies that don't evolve their metrics end up optimizing for the wrong thing at exactly the moment it's most costly to do so.

Vanity Metrics vs. Clarity Metrics

There's a useful distinction I keep coming back to. Vanity metrics make you feel good about the work. Clarity metrics tell you whether the work is actually producing anything.

• Page views versus conversion
• Emails sent versus pipeline actually created
• Features shipped versus whether anyone uses them six months later

Vanity metrics are addictive because they're usually going up. Clarity metrics are uncomfortable because they force the question: is any of this working?

When teams only measure activity, you also lose something culturally important — the ability to have honest conversations about what's not working. If the metric is "we shipped 12 features" and you shipped 12, everyone gets to feel successful regardless of whether any of those features moved the needle. That makes course-correction very hard.

The Fix Is Simpler Than You Think

Attach every metric to a decision.

Before you add a number to your dashboard, ask: "What would we do differently if this went up versus down?" If the answer is "nothing" or "we'd investigate further" — it's not a metric. It's a monitoring signal.

A real metric tells the team exactly what to do when it moves. Gross margin compresses — look at COGS, and right now that increasingly means AI infrastructure spend. Net revenue retention dips below 100% — you're in contraction, go talk to customers. Revenue per employee drops — look at hiring pace relative to growth.

The metric doesn't just tell you something happened. It tells you where to look and what to do, immediately.

The second piece: the metric needs to be visible to the people who can actually influence it. Leadership obsessing over a number that individual contributors don't even know exists is completely backwards. The most effective leaders I've seen take one metric — just one — and make it everyone's problem in the best possible way. On the wall, in every meeting, until everyone can tell you the number without looking it up. People make better decisions because they have a shared definition of what they're working toward. You stop forcing alignment because it just happens.

The shift is simple to describe and hard to do: be willing to make a bet. Say, "This is the number that matters most right now. We're going to be ruthless about it, and everything else is secondary."

That implies trade-offs. It implies you might be wrong about what matters. But that clarity is what frees a team to move fast.

Ambiguity is the enemy of execution.